Thu May 20 2021

AI risk assessment and management in high-risk applications

LinkedIn Live
Thu May 20 202114:01
LinkedIn Live
Join event

About this event:

The May edition of the Responsible AI webinar series revolves around the topic "AI risk assessment and management in high-risk applications".

Our first guest speaker will be Alexander Boer, a senior manager of trusted analytics at KPMG. He will talk about the role of trust in the use of AI applications. Especially in high-risk environments, such as finance, the accurate and robust performance of AI systems must be guaranteed; Alexander will show us how AI and the surrounding organization can be designed and operated in a way that safeguards potential risks. Examples include the use of transparent, interpretable, and explainable AI, and auditing mechanisms to test for regulatory compliance. This approach works well in domains where there is a societal risk of organizations violating rules and regulations (intentionally or accidentally), such as fraud.

Our second guest speaker will be Mark Roboff, a general manager at DXC Technology in Virginia, USA, working in the domain of Aerospace and Defense - also a high-risk environment. Mark will provide us with a slightly different take on risk assessment and management by focusing on simulation-based testing. This approach requires an AI system to perform numerous times across slight variations of scenarios in a simulated environment to test its behavior across many different situations and parameters. Simulation-based testing is particularly suited for domains such as aviation, whereby it is expected that a company is not only legally obligated but also innately dedicated, to prevent its systems from misbehaving or behaving erratically under certain unforeseen circumstances.

Program:

  • 16:00: opening by Marieke Peeters
  • 16:05: talk by Alexander Boer
  • 16:25: Q&A
  • 16:30: talk by Mark Roboff
  • 16:55: Q&A
  • 17:00: closing by Marieke